Does Parents’ Spatial Multimodal Input Structure Infants’ Attention During Puzzle Play?

• In Event: Why Parent-Infant Interaction Matters: Impacts of Parent’s Multimodal Inputs on Infant’s Perceptual Experiences

Abstract

Spatial skills are an integral part of daily activities, such as navigation and finding our keys. Spatial abilities have their origins in infancy and develop across childhood (e.g., Uttal et al., 2013). Parents’ spatial language input influences children’s performance on spatial tasks concurrently and longitudinally during the preschool years (e.g., Pruden et al., 2011; Ralph et al., 2020). Although exposure to language-particular ways of encoding spatial information shapes how infants parse and process space (e.g., Casasola & Ahn, 2018), it is unknown whether parents’ spatial language influences infants’ distribution of attention during shared play. Previous research has shown that infants’ attention to objects is influenced by parents’ speech and gestures (e.g., Deák et al., 2014; Yu & Smith, 2013; 2017) and by multimodal parent behaviors (touching an object while talking) within joint attention (Suarez-Rivera et al., 2019). Here, we examine how parents’ spatial multimodal input, both their spatial terms and deictic gestures (e.g., pointing), structures infants’ attention during puzzle play.

We recorded 3-min puzzle play sessions with 53 8- to 11-month-old infants (Mage=9.15 months, SD=.80) and their parents. Both play partners were fitted with Positive Science head-mounted eye trackers, allowing us to record eye gaze as the dyads played with an age-appropriate puzzle. We coded by looking at the play partner’s face, partner’s hand, and puzzle, and parents’ speech was transcribed. We will code parents’ use of spatial terms using the System for Analyzing Children’s Language about Space (Cannon et al., 2007), which categorizes English spatial terms by size, shape, features, locations, orientation or transformation of entities, spatial relations between entities, or deictic words (e.g., small, square, edge, next to, above, turn, there). Parent object labeling will also be coded to provide a comparison. We will code parents’ deictic gestures, hand movements, such as pointing, that indicate a referent in the immediate environment (McNeill, 1992). We will use onsets and offsets of behaviors for time-series analyses.

Our planned analyses will address three predictions. First, we predict that overall parents’ spatial terms and deictic gestures will be associated with infants’ looking time to both the puzzle pieces and where they should be inserted, suggesting that even in infancy parental spatial input shapes infants’ attention. We will conduct hierarchical linear regressions predicting infants’ looking time to puzzle from parents’ multimodal behaviors (including a contrast between spatial terms and object labels), controlling for parents’ total utterances and infants’ age. Second, we predict that parents’ multimodal spatial input will temporally precede infants’ looking to the puzzle, indicating that not only are infants’ behaviors related to parental input, but spatial input specifically induces shifts in infants’ attention. We will test this prediction with regression analyses with lagged predictors demonstrating how these relations between infants’ looking and parents’ multimodal spatial input unfold over time. Finally, we predict that these relations will be strongest when parental input is multimodal, when parents provide spatial terms and deictic gestures simultaneously, implying that coordinated speech and gesture have complementary roles in visual attention.

Authors

• Erim Kızıldere, University of California - Davis
  Presenting Author

• Christian Nelson, University of California Davis

• Marianella Casasola, Cornell University

• Lisa M Oakes, University of California - Davis